Finding experiments

To use incense we first have to instantiate an experiment loader that will enable us to query the database for specific runs.

targets_type iteration autoencoder_type batch_size artifacts
exp_id
58 Mnist False Over_dim_tied_iteration 256 {'history_autoencoder_iteration': Artifact(nam...
59 Mnist False Over_dim_tied_iteration 128 {'history_autoencoder_iteration': Artifact(nam...
60 Mnist False Over_dim_tied_iteration 64 {'history_autoencoder_iteration': Artifact(nam...
61 Mnist False Over_dim_tied_iteration 32 {'history_autoencoder_iteration': Artifact(nam...
62 10_Targets False Over_dim_tied_iteration 256 {'history_autoencoder_iteration': Artifact(nam...
63 10_Targets False Over_dim_tied_iteration 128 {'history_autoencoder_iteration': Artifact(nam...
64 10_Targets False Over_dim_tied_iteration 64 {'history_autoencoder_iteration': Artifact(nam...
65 10_Targets False Over_dim_tied_iteration 32 {'history_autoencoder_iteration': Artifact(nam...
targets_type iteration autoencoder_type batch_size artifacts sort
exp_id
62 10_Targets False Over_dim_tied_iteration 256 {'history_autoencoder_iteration': Artifact(nam... 0
63 10_Targets False Over_dim_tied_iteration 128 {'history_autoencoder_iteration': Artifact(nam... 1
64 10_Targets False Over_dim_tied_iteration 64 {'history_autoencoder_iteration': Artifact(nam... 2
65 10_Targets False Over_dim_tied_iteration 32 {'history_autoencoder_iteration': Artifact(nam... 3
58 Mnist False Over_dim_tied_iteration 256 {'history_autoencoder_iteration': Artifact(nam... 4
59 Mnist False Over_dim_tied_iteration 128 {'history_autoencoder_iteration': Artifact(nam... 5
60 Mnist False Over_dim_tied_iteration 64 {'history_autoencoder_iteration': Artifact(nam... 6
61 Mnist False Over_dim_tied_iteration 32 {'history_autoencoder_iteration': Artifact(nam... 7

Red best overall, and also best of subset. Bes means for accuracy max, rest min. Green best of subset.

predictions_df_0
Accuracy over iterations evaluations_feature_classifier
Over_dim_tied_iteration 256 10_Targets Over_dim_tied_iteration 128 10_Targets Over_dim_tied_iteration 64 10_Targets Over_dim_tied_iteration 32 10_Targets Over_dim_tied_iteration 256 Mnist Over_dim_tied_iteration 128 Mnist Over_dim_tied_iteration 64 Mnist Over_dim_tied_iteration 32 Mnist
0 0.9751 0.9602 0.5022 0.8382 0.9777 0.9781 0.976 0.3537
1 0.902 0.8754 0.1135 0.5707 0.9758 0.977 0.9758 0.1274
2 0.8328 0.8082 0.1135 0.5289 0.9702 0.9733 0.9732 0.1039
3 0.7978 0.7682 0.1135 0.5163 0.9637 0.9667 0.9673 0.1028
4 0.7785 0.7394 0.1135 0.5112 0.9574 0.9577 0.9611 0.1028
5 0.769 0.7212 0.1135 0.5098 0.9476 0.948 0.9528 0.1028
6 0.7617 0.7068 0.1135 0.5083 0.936 0.9332 0.9395 0.1028
7 0.758 0.6969 0.1135 0.5079 0.92 0.915 0.9253 0.1028
Loss over iterations autoencoder
Over_dim_tied_iteration 256 10_Targets Over_dim_tied_iteration 128 10_Targets Over_dim_tied_iteration 64 10_Targets Over_dim_tied_iteration 32 10_Targets Over_dim_tied_iteration 256 Mnist Over_dim_tied_iteration 128 Mnist Over_dim_tied_iteration 64 Mnist Over_dim_tied_iteration 32 Mnist
0 0.365515 0.344711 0.455832 0.299053 0.0157922 0.0158386 0.0141918 0.381717
1 0.383135 29265.5 0.455832 162675 0.0267629 0.0275521 0.0228369 0.382225
2 0.393784 1.41281e+13 0.455832 3.33656e+12 0.0408471 0.0424907 0.0340128 0.382359
3 0.399944 6.82077e+21 0.455832 6.84377e+19 0.0568863 4.56198e+07 0.0468889 0.382421
4 0.40428 3.29294e+30 0.455832 1.40376e+27 0.0741922 2.28454e+24 35218.1 0.382439
5 0.407509 inf 0.455832 inf 7.43304e+07 inf 2.11025e+21 0.382443
6 0.409831 inf 0.455832 inf 3.47823e+23 inf inf 0.382444
7 0.411346 inf 0.455832 inf inf inf inf 0.382445
MAE over iterations autoencoder
Over_dim_tied_iteration 256 10_Targets Over_dim_tied_iteration 128 10_Targets Over_dim_tied_iteration 64 10_Targets Over_dim_tied_iteration 32 10_Targets Over_dim_tied_iteration 256 Mnist Over_dim_tied_iteration 128 Mnist Over_dim_tied_iteration 64 Mnist Over_dim_tied_iteration 32 Mnist
0 0.258537 0.25949 0.265029 0.25146 0.0418313 0.0421456 0.0404086 0.264885
1 0.262569 1.75583 0.265029 4.08223 0.0533651 0.0544632 0.0503333 0.2652
2 0.264747 32754.5 0.265029 17316 0.0654345 0.0671523 0.0608999 0.265247
3 0.266198 7.19687e+08 0.265029 7.8422e+07 0.0772814 69.7403 0.0713959 0.265267
4 0.26706 1.58132e+13 0.265029 3.5517e+11 0.08871 1.55883e+10 2.42887 0.265273
5 0.267646 3.47451e+17 0.265029 1.60855e+15 184.664 3.48837e+18 5.74526e+08 0.265274
6 0.26807 7.63429e+21 0.265029 7.28505e+18 1.26258e+10 7.80632e+26 1.40645e+17 0.265274
7 0.268339 1.67743e+26 0.265029 3.29937e+22 8.63687e+17 inf 3.44299e+25 0.265275 
predictions_df_10
Accuracy over iterations evaluations_feature_classifier
Over_dim_tied_iteration 256 10_Targets Over_dim_tied_iteration 128 10_Targets Over_dim_tied_iteration 64 10_Targets Over_dim_tied_iteration 32 10_Targets Over_dim_tied_iteration 256 Mnist Over_dim_tied_iteration 128 Mnist Over_dim_tied_iteration 64 Mnist Over_dim_tied_iteration 32 Mnist
0 0.9646 0.9512 0.4513 0.8199 0.9516 0.9557 0.9548 0.3043
1 0.8851 0.8603 0.1135 0.5655 0.9596 0.9657 0.9612 0.1161
2 0.8078 0.791 0.1135 0.5198 0.9578 0.9631 0.9588 0.1032
3 0.7755 0.7519 0.1135 0.5035 0.9498 0.9551 0.9544 0.1028
4 0.7555 0.7218 0.1135 0.4978 0.9406 0.944 0.9458 0.1028
5 0.7442 0.7023 0.1135 0.4967 0.9283 0.9317 0.9341 0.1028
6 0.7376 0.6904 0.1135 0.4959 0.9095 0.9137 0.9174 0.1028
7 0.7331 0.6816 0.1135 0.4956 0.8913 0.8937 0.8983 0.1028
Loss over iterations autoencoder
Over_dim_tied_iteration 256 10_Targets Over_dim_tied_iteration 128 10_Targets Over_dim_tied_iteration 64 10_Targets Over_dim_tied_iteration 32 10_Targets Over_dim_tied_iteration 256 Mnist Over_dim_tied_iteration 128 Mnist Over_dim_tied_iteration 64 Mnist Over_dim_tied_iteration 32 Mnist
0 0.36237 0.341752 0.455832 0.28745 0.0423389 7.8254e+06 5.28285e+07 0.381441
1 0.3818 0.369189 0.455832 0.344956 0.0498952 3.91871e+23 3.16585e+24 0.382245
2 0.394055 0.386584 0.455832 37766.9 0.0625914 inf inf 0.382367
3 0.400722 0.401297 0.455832 7.74555e+11 0.0779077 inf inf 0.382424
4 0.405389 0.414306 0.455832 1.58873e+19 1.8792e+08 inf inf 0.382439
5 0.408513 0.423468 0.455832 3.25871e+26 8.79355e+23 nan nan 0.382443
6 0.410958 0.430074 0.455832 inf inf nan nan 0.382444
7 0.412752 0.434429 0.455832 inf inf nan nan 0.382445
MAE over iterations autoencoder
Over_dim_tied_iteration 256 10_Targets Over_dim_tied_iteration 128 10_Targets Over_dim_tied_iteration 64 10_Targets Over_dim_tied_iteration 32 10_Targets Over_dim_tied_iteration 256 Mnist Over_dim_tied_iteration 128 Mnist Over_dim_tied_iteration 64 Mnist Over_dim_tied_iteration 32 Mnist
0 0.258278 0.259856 0.265029 0.252088 0.0697001 63.3859 279.488 0.264782
1 0.262572 0.265991 0.265029 0.26107 0.0737914 1.41672e+10 6.84034e+10 0.265172
2 0.265109 0.271233 0.265029 1.96198 0.0822621 3.17035e+18 1.67452e+19 0.265238
3 0.266622 0.275797 0.265029 7673.86 0.091958 7.09464e+26 4.09924e+27 0.265265
4 0.267631 0.27967 0.265029 3.47534e+07 156.684 inf inf 0.265272
5 0.26822 0.282567 0.265029 1.57397e+11 1.07114e+10 nan nan 0.265274
6 0.268671 0.284722 0.265029 7.12843e+14 7.32726e+17 nan nan 0.265274
7 0.269024 0.286244 0.265029 3.22844e+18 5.01231e+25 nan nan 0.265275 
predictions_df_20
Accuracy over iterations evaluations_feature_classifier
Over_dim_tied_iteration 256 10_Targets Over_dim_tied_iteration 128 10_Targets Over_dim_tied_iteration 64 10_Targets Over_dim_tied_iteration 32 10_Targets Over_dim_tied_iteration 256 Mnist Over_dim_tied_iteration 128 Mnist Over_dim_tied_iteration 64 Mnist Over_dim_tied_iteration 32 Mnist
0 0.9528 0.9404 0.355 0.7929 0.9148 0.916 0.919 0.2611
1 0.8598 0.8416 0.1135 0.5634 0.9348 0.9416 0.9354 0.109
2 0.7796 0.774 0.1135 0.5066 0.9308 0.9398 0.9347 0.1029
3 0.7458 0.7335 0.1135 0.4882 0.9196 0.933 0.9272 0.1028
4 0.7292 0.7042 0.1135 0.4832 0.906 0.9201 0.9131 0.1028
5 0.7168 0.6887 0.1135 0.4809 0.8891 0.9001 0.8942 0.1028
6 0.7099 0.6759 0.1135 0.4802 0.8677 0.8817 0.8751 0.1028
7 0.7049 0.6685 0.1135 0.48 0.8438 0.8591 0.8544 0.1028
Loss over iterations autoencoder
Over_dim_tied_iteration 256 10_Targets Over_dim_tied_iteration 128 10_Targets Over_dim_tied_iteration 64 10_Targets Over_dim_tied_iteration 32 10_Targets Over_dim_tied_iteration 256 Mnist Over_dim_tied_iteration 128 Mnist Over_dim_tied_iteration 64 Mnist Over_dim_tied_iteration 32 Mnist
0 0.359457 0.338906 0.455832 0.28566 0.0659356 1.17554e+08 3.607e+08 0.381203
1 0.381086 0.368372 0.455832 0.342205 0.0720645 5.88677e+24 2.16157e+25 0.382256
2 0.394452 0.387029 0.455832 0.382495 9.61913e+07 inf inf 0.382374
3 0.401379 0.403141 0.455832 0.404541 4.50119e+23 inf inf 0.382425
4 0.406072 0.417865 0.455832 0.415999 inf inf nan 0.38244
5 0.409475 0.428316 0.455832 0.421818 inf nan nan 0.382443
6 0.411821 0.435205 0.455832 0.424462 inf nan nan 0.382444
7 0.413506 0.439611 0.455832 0.42568 nan nan nan 0.382445
MAE over iterations autoencoder
Over_dim_tied_iteration 256 10_Targets Over_dim_tied_iteration 128 10_Targets Over_dim_tied_iteration 64 10_Targets Over_dim_tied_iteration 32 10_Targets Over_dim_tied_iteration 256 Mnist Over_dim_tied_iteration 128 Mnist Over_dim_tied_iteration 64 Mnist Over_dim_tied_iteration 32 Mnist
0 0.258337 0.260514 0.265029 0.253837 0.0891158 475.75 1120.34 0.264649
1 0.262999 0.267079 0.265029 0.262599 0.0905341 1.06435e+11 2.74241e+11 0.265137
2 0.265636 0.272725 0.265029 0.270225 120.947 2.38182e+19 6.71344e+19 0.26523
3 0.267001 0.277685 0.265029 0.274898 8.26698e+09 5.33005e+27 1.64345e+28 0.265263
4 0.267922 0.282041 0.265029 0.277418 5.65513e+17 inf nan 0.265271
5 0.268517 0.285193 0.265029 0.278654 3.86846e+25 nan nan 0.265274
6 0.268887 0.28735 0.265029 0.279223 inf nan nan 0.265274
7 0.26921 0.288814 0.265029 0.279467 nan nan nan 0.265275 
predictions_df_30
Accuracy over iterations evaluations_feature_classifier
Over_dim_tied_iteration 256 10_Targets Over_dim_tied_iteration 128 10_Targets Over_dim_tied_iteration 64 10_Targets Over_dim_tied_iteration 32 10_Targets Over_dim_tied_iteration 256 Mnist Over_dim_tied_iteration 128 Mnist Over_dim_tied_iteration 64 Mnist Over_dim_tied_iteration 32 Mnist
0 0.9307 0.9116 0.2649 0.7527 0.8689 0.8663 0.8812 0.2199
1 0.826 0.8137 0.1135 0.545 0.8978 0.9068 0.9065 0.1063
2 0.7485 0.7446 0.1135 0.4916 0.8942 0.9058 0.905 0.103
3 0.7196 0.7078 0.1135 0.4696 0.881 0.8967 0.8958 0.1028
4 0.7031 0.6798 0.1135 0.4621 0.8632 0.8823 0.8799 0.1028
5 0.6911 0.6629 0.1135 0.4601 0.8401 0.8632 0.8599 0.1028
6 0.684 0.652 0.1135 0.4599 0.8145 0.8415 0.8374 0.1028
7 0.6787 0.6455 0.1135 0.46 0.7898 0.818 0.8139 0.1028
Loss over iterations autoencoder
Over_dim_tied_iteration 256 10_Targets Over_dim_tied_iteration 128 10_Targets Over_dim_tied_iteration 64 10_Targets Over_dim_tied_iteration 32 10_Targets Over_dim_tied_iteration 256 Mnist Over_dim_tied_iteration 128 Mnist Over_dim_tied_iteration 64 Mnist Over_dim_tied_iteration 32 Mnist
0 0.356639 0.336023 0.455832 0.284259 4.21162e+06 3.22352e+08 7.00259e+08 0.380676
1 0.382153 0.368424 0.455832 0.342949 1.97078e+22 1.61426e+25 4.19645e+25 0.382182
2 0.39611 0.389777 0.455832 0.385501 inf inf inf 0.382354
3 0.403344 0.407545 0.455832 0.409722 inf inf inf 0.382421
4 0.408102 0.423431 0.455832 0.422368 inf nan nan 0.382439
5 0.411636 0.434447 0.455832 0.428781 nan nan nan 0.382443
6 0.413949 0.4415 0.455832 0.431519 nan nan nan 0.382444
7 0.415559 0.445994 0.455832 0.432734 nan nan nan 0.382445
MAE over iterations autoencoder
Over_dim_tied_iteration 256 10_Targets Over_dim_tied_iteration 128 10_Targets Over_dim_tied_iteration 64 10_Targets Over_dim_tied_iteration 32 10_Targets Over_dim_tied_iteration 256 Mnist Over_dim_tied_iteration 128 Mnist Over_dim_tied_iteration 64 Mnist Over_dim_tied_iteration 32 Mnist
0 0.259134 0.261901 0.265029 0.255998 56.2033 1053.86 1837.66 0.264363
1 0.264191 0.269075 0.265029 0.264821 3.83789e+09 2.35793e+11 4.49837e+11 0.265057
2 0.266636 0.27537 0.265029 0.273113 2.62536e+17 5.2766e+19 1.10121e+20 0.265209
3 0.268021 0.280764 0.265029 0.278283 1.79591e+25 1.1808e+28 2.69576e+28 0.265258
4 0.268939 0.285382 0.265029 0.281038 inf nan nan 0.26527
5 0.269575 0.288599 0.265029 0.282385 nan nan nan 0.265273
6 0.26996 0.290792 0.265029 0.282928 nan nan nan 0.265274
7 0.270223 0.292267 0.265029 0.283154 nan nan nan 0.265274 
predictions_df_40
Accuracy over iterations evaluations_feature_classifier
Over_dim_tied_iteration 256 10_Targets Over_dim_tied_iteration 128 10_Targets Over_dim_tied_iteration 64 10_Targets Over_dim_tied_iteration 32 10_Targets Over_dim_tied_iteration 256 Mnist Over_dim_tied_iteration 128 Mnist Over_dim_tied_iteration 64 Mnist Over_dim_tied_iteration 32 Mnist
0 0.8902 0.8749 0.2143 0.703 0.8192 0.8142 0.8248 0.184
1 0.7871 0.774 0.1135 0.529 0.85 0.8549 0.854 0.1053
2 0.7104 0.71 0.1135 0.4747 0.846 0.8597 0.8552 0.1029
3 0.6767 0.6753 0.1135 0.4573 0.8322 0.8526 0.8444 0.1029
4 0.6578 0.648 0.1135 0.4499 0.8129 0.838 0.8277 0.1029
5 0.6451 0.6311 0.1135 0.4468 0.7902 0.817 0.8072 0.1029
6 0.6396 0.6204 0.1135 0.4464 0.7655 0.7925 0.7847 0.1029
7 0.6354 0.6134 0.1135 0.4463 0.7358 0.765 0.7602 0.1029
Loss over iterations autoencoder
Over_dim_tied_iteration 256 10_Targets Over_dim_tied_iteration 128 10_Targets Over_dim_tied_iteration 64 10_Targets Over_dim_tied_iteration 32 10_Targets Over_dim_tied_iteration 256 Mnist Over_dim_tied_iteration 128 Mnist Over_dim_tied_iteration 64 Mnist Over_dim_tied_iteration 32 Mnist
0 0.352247 0.333506 0.455832 0.28391 2.23681e+07 7.50971e+08 8.40472e+08 0.383479
1 0.381138 0.369243 0.455832 0.345825 1.04669e+23 3.76068e+25 5.03671e+25 8.83669
2 0.396389 0.393616 0.455832 0.39216 inf inf inf 8.83691
3 0.404912 0.413641 0.455832 0.417924 inf inf inf 8.83699
4 0.410318 0.430681 0.455832 0.431377 inf nan nan 8.83701
5 0.413839 0.442381 0.455832 0.438145 nan nan nan 8.83701
6 0.416148 0.449849 0.455832 0.441277 nan nan nan 8.83702
7 0.417803 0.454675 0.455832 0.442651 nan nan nan 8.83702
MAE over iterations autoencoder
Over_dim_tied_iteration 256 10_Targets Over_dim_tied_iteration 128 10_Targets Over_dim_tied_iteration 64 10_Targets Over_dim_tied_iteration 32 10_Targets Over_dim_tied_iteration 256 Mnist Over_dim_tied_iteration 128 Mnist Over_dim_tied_iteration 64 Mnist Over_dim_tied_iteration 32 Mnist
0 0.259634 0.263801 0.265029 0.258862 159.226 1841.29 2046.98 0.264629
1 0.264975 0.271778 0.265029 0.268088 1.08848e+10 4.11998e+11 5.01075e+11 0.300326
2 0.267553 0.278895 0.265029 0.277158 7.44592e+17 9.21972e+19 1.22664e+20 0.30054
3 0.269104 0.284811 0.265029 0.282745 5.09347e+25 2.0632e+28 3.00281e+28 0.300606
4 0.270094 0.289681 0.265029 0.28577 inf nan nan 0.300624
5 0.270682 0.293102 0.265029 0.28724 nan nan nan 0.300628
6 0.271038 0.295354 0.265029 0.287879 nan nan nan 0.300629
7 0.271325 0.296904 0.265029 0.288151 nan nan nan 0.300629 
predictions_df_50
Accuracy over iterations evaluations_feature_classifier
Over_dim_tied_iteration 256 10_Targets Over_dim_tied_iteration 128 10_Targets Over_dim_tied_iteration 64 10_Targets Over_dim_tied_iteration 32 10_Targets Over_dim_tied_iteration 256 Mnist Over_dim_tied_iteration 128 Mnist Over_dim_tied_iteration 64 Mnist Over_dim_tied_iteration 32 Mnist
0 0.8446 0.8315 0.1916 0.6435 0.762 0.7483 0.7689 0.1408
1 0.7456 0.718 0.1135 0.5062 0.7905 0.798 0.803 0.1064
2 0.6764 0.6505 0.1135 0.4524 0.7901 0.8043 0.8049 0.1031
3 0.6486 0.6164 0.1135 0.4315 0.7788 0.7927 0.7912 0.103
4 0.6327 0.5946 0.1135 0.4255 0.7521 0.7719 0.7726 0.103
5 0.6216 0.5832 0.1135 0.423 0.725 0.7506 0.753 0.103
6 0.6149 0.5754 0.1135 0.4221 0.6976 0.725 0.7286 0.103
7 0.6099 0.5692 0.1135 0.4219 0.6723 0.7024 0.7043 0.103
Loss over iterations autoencoder
Over_dim_tied_iteration 256 10_Targets Over_dim_tied_iteration 128 10_Targets Over_dim_tied_iteration 64 10_Targets Over_dim_tied_iteration 32 10_Targets Over_dim_tied_iteration 256 Mnist Over_dim_tied_iteration 128 Mnist Over_dim_tied_iteration 64 Mnist Over_dim_tied_iteration 32 Mnist
0 0.348702 0.330399 0.455832 0.283863 2.43769e+08 3.132e+09 2.75913e+09 0.421153
1 0.380188 0.370205 0.455832 0.348805 1.14069e+24 1.56844e+26 1.65347e+26 21.5301
2 0.397133 0.399206 0.455832 0.398787 inf inf inf 21.5304
3 0.405673 0.423216 0.455832 0.426219 inf inf inf 21.5305
4 0.410899 0.442688 0.455832 0.440156 inf nan nan 21.5305
5 0.414661 0.455425 0.455832 0.447267 nan nan nan 21.5305
6 0.417293 0.462817 0.455832 0.45064 nan nan nan 21.5305
7 0.419109 0.467238 0.455832 0.45208 nan nan nan 21.5305
MAE over iterations autoencoder
Over_dim_tied_iteration 256 10_Targets Over_dim_tied_iteration 128 10_Targets Over_dim_tied_iteration 64 10_Targets Over_dim_tied_iteration 32 10_Targets Over_dim_tied_iteration 256 Mnist Over_dim_tied_iteration 128 Mnist Over_dim_tied_iteration 64 Mnist Over_dim_tied_iteration 32 Mnist
0 0.260999 0.266667 0.265029 0.262074 935.544 4908.02 4751.9 0.266758
1 0.265838 0.275847 0.265029 0.271626 6.39918e+10 1.09825e+12 1.16324e+12 0.353396
2 0.268396 0.284265 0.265029 0.281432 4.37744e+18 2.45767e+20 2.84762e+20 0.353661
3 0.269674 0.29115 0.265029 0.287471 2.99444e+26 5.4998e+28 6.971e+28 0.353746
4 0.270451 0.296502 0.265029 0.290606 inf nan nan 0.353769
5 0.271091 0.300013 0.265029 0.292153 nan nan nan 0.353774
6 0.271571 0.302154 0.265029 0.292831 nan nan nan 0.353776
7 0.271871 0.303532 0.265029 0.293116 nan nan nan 0.353776 
predictions_df_60
Accuracy over iterations evaluations_feature_classifier
Over_dim_tied_iteration 256 10_Targets Over_dim_tied_iteration 128 10_Targets Over_dim_tied_iteration 64 10_Targets Over_dim_tied_iteration 32 10_Targets Over_dim_tied_iteration 256 Mnist Over_dim_tied_iteration 128 Mnist Over_dim_tied_iteration 64 Mnist Over_dim_tied_iteration 32 Mnist
0 0.7723 0.7634 0.1726 0.5805 0.704 0.6813 0.7056 0.1183
1 0.6829 0.6588 0.1135 0.4729 0.7283 0.7225 0.7424 0.1054
2 0.6249 0.5971 0.1135 0.4284 0.7214 0.7306 0.7448 0.103
3 0.5986 0.5651 0.1135 0.4084 0.7055 0.7207 0.7289 0.103
4 0.5853 0.5455 0.1135 0.4027 0.6841 0.7028 0.7077 0.103
5 0.5729 0.5351 0.1135 0.4003 0.6588 0.6822 0.6862 0.103
6 0.566 0.5283 0.1135 0.3992 0.6302 0.66 0.666 0.103
7 0.563 0.5242 0.1135 0.3992 0.605 0.6353 0.6422 0.103
Loss over iterations autoencoder
Over_dim_tied_iteration 256 10_Targets Over_dim_tied_iteration 128 10_Targets Over_dim_tied_iteration 64 10_Targets Over_dim_tied_iteration 32 10_Targets Over_dim_tied_iteration 256 Mnist Over_dim_tied_iteration 128 Mnist Over_dim_tied_iteration 64 Mnist Over_dim_tied_iteration 32 Mnist
0 0.346268 0.330528 0.45586 0.285116 4.10346e+08 3.7645e+09 3.35784e+09 0.639216
1 0.38144 0.374805 0.455832 0.354166 1.92018e+24 1.88518e+26 2.01226e+26 55.375
2 0.399868 0.406582 0.455832 0.40825 inf inf inf 55.3751
3 0.40871 0.433265 0.455832 0.437251 inf inf inf 55.3752
4 0.414274 0.454806 0.455832 0.45187 inf nan nan 55.3752
5 0.418373 0.468483 0.455832 0.458995 nan nan nan 55.3752
6 0.421099 0.476399 0.455832 0.462409 nan nan nan 55.3752
7 0.422988 0.480929 0.455832 0.463859 nan nan nan 55.3752
MAE over iterations autoencoder
Over_dim_tied_iteration 256 10_Targets Over_dim_tied_iteration 128 10_Targets Over_dim_tied_iteration 64 10_Targets Over_dim_tied_iteration 32 10_Targets Over_dim_tied_iteration 256 Mnist Over_dim_tied_iteration 128 Mnist Over_dim_tied_iteration 64 Mnist Over_dim_tied_iteration 32 Mnist
0 0.263403 0.270365 0.265031 0.266306 1690.16 5984.37 5459.51 0.277732
1 0.267936 0.280857 0.265029 0.276591 1.15614e+11 1.3391e+12 1.33646e+12 0.495049
2 0.270106 0.290024 0.265029 0.287057 7.90869e+18 2.99665e+20 3.27167e+20 0.495305
3 0.271326 0.297545 0.265029 0.293343 5.41004e+26 6.70594e+28 8.00907e+28 0.495398
4 0.272211 0.303374 0.265029 0.296598 inf nan nan 0.495424
5 0.272842 0.3071 0.265029 0.298169 nan nan nan 0.49543
6 0.273261 0.309381 0.265029 0.298891 nan nan nan 0.495432
7 0.273606 0.310766 0.265029 0.299172 nan nan nan 0.495432 
predictions_df_70
Accuracy over iterations evaluations_feature_classifier
Over_dim_tied_iteration 256 10_Targets Over_dim_tied_iteration 128 10_Targets Over_dim_tied_iteration 64 10_Targets Over_dim_tied_iteration 32 10_Targets Over_dim_tied_iteration 256 Mnist Over_dim_tied_iteration 128 Mnist Over_dim_tied_iteration 64 Mnist Over_dim_tied_iteration 32 Mnist
0 0.6872 0.6905 0.166 0.5129 0.6275 0.5978 0.6352 0.1025
1 0.6225 0.5899 0.1135 0.4408 0.6494 0.6308 0.6627 0.1087
2 0.5732 0.5301 0.1135 0.3958 0.6443 0.6384 0.6618 0.1034
3 0.5519 0.5007 0.1135 0.3793 0.6244 0.6274 0.6496 0.1034
4 0.5386 0.4835 0.1135 0.3733 0.6025 0.6074 0.6339 0.1033
5 0.528 0.4741 0.1135 0.3714 0.5837 0.5872 0.6141 0.1033
6 0.5216 0.4688 0.1135 0.3707 0.5582 0.5692 0.5935 0.1033
7 0.5175 0.4647 0.1135 0.3702 0.5384 0.5521 0.5733 0.1033
Loss over iterations autoencoder
Over_dim_tied_iteration 256 10_Targets Over_dim_tied_iteration 128 10_Targets Over_dim_tied_iteration 64 10_Targets Over_dim_tied_iteration 32 10_Targets Over_dim_tied_iteration 256 Mnist Over_dim_tied_iteration 128 Mnist Over_dim_tied_iteration 64 Mnist Over_dim_tied_iteration 32 Mnist
0 0.345311 0.331163 0.455879 0.287811 1.10533e+09 8.74675e+09 5.97337e+09 1.68987
1 0.38267 0.381589 0.455832 0.361305 5.1723e+24 4.38018e+26 3.57968e+26 178.328
2 0.401484 0.417522 0.455832 0.419539 inf inf inf 178.066
3 0.410695 0.445576 0.455832 0.450089 inf inf inf 178.066
4 0.416689 0.467899 0.455832 0.464693 inf nan nan 178.066
5 0.421133 0.482281 0.455832 0.471655 nan nan nan 178.066
6 0.424612 0.490507 0.455832 0.475162 nan nan nan 178.066
7 0.426578 0.494963 0.455832 0.476928 nan nan nan 178.066
MAE over iterations autoencoder
Over_dim_tied_iteration 256 10_Targets Over_dim_tied_iteration 128 10_Targets Over_dim_tied_iteration 64 10_Targets Over_dim_tied_iteration 32 10_Targets Over_dim_tied_iteration 256 Mnist Over_dim_tied_iteration 128 Mnist Over_dim_tied_iteration 64 Mnist Over_dim_tied_iteration 32 Mnist
0 0.266741 0.275313 0.265031 0.271139 3621.85 11787.1 8252.29 0.310971
1 0.270541 0.28689 0.265029 0.282079 2.47758e+11 2.63761e+12 2.02013e+12 1.0094
2 0.272201 0.296816 0.265029 0.293328 1.69482e+19 5.90247e+20 4.9453e+20 1.00902
3 0.273173 0.304589 0.265029 0.300047 1.15936e+27 1.32086e+29 1.21061e+29 1.0091
4 0.274019 0.310635 0.265029 0.303301 inf nan nan 1.00913
5 0.274695 0.314567 0.265029 0.304806 nan nan nan 1.00913
6 0.275201 0.316937 0.265029 0.30553 nan nan nan 1.00914
7 0.275476 0.318265 0.265029 0.305872 nan nan nan 1.00914 
predictions_df_80
Accuracy over iterations evaluations_feature_classifier
Over_dim_tied_iteration 256 10_Targets Over_dim_tied_iteration 128 10_Targets Over_dim_tied_iteration 64 10_Targets Over_dim_tied_iteration 32 10_Targets Over_dim_tied_iteration 256 Mnist Over_dim_tied_iteration 128 Mnist Over_dim_tied_iteration 64 Mnist Over_dim_tied_iteration 32 Mnist
0 0.5941 0.614 0.1561 0.4474 0.5578 0.5126 0.554 0.0982
1 0.5529 0.5198 0.1135 0.3905 0.5777 0.5473 0.5822 0.1093
2 0.5116 0.4669 0.1135 0.3597 0.5676 0.5533 0.5822 0.1034
3 0.4933 0.4402 0.1135 0.3448 0.5536 0.5459 0.5685 0.1029
4 0.4811 0.4262 0.1135 0.3407 0.5305 0.5323 0.5544 0.1029
5 0.4696 0.4178 0.1135 0.3387 0.5105 0.509 0.535 0.1029
6 0.4641 0.4135 0.1135 0.3386 0.49 0.4912 0.5162 0.1029
7 0.4607 0.4103 0.1135 0.3386 0.4721 0.4758 0.4979 0.1029
Loss over iterations autoencoder
Over_dim_tied_iteration 256 10_Targets Over_dim_tied_iteration 128 10_Targets Over_dim_tied_iteration 64 10_Targets Over_dim_tied_iteration 32 10_Targets Over_dim_tied_iteration 256 Mnist Over_dim_tied_iteration 128 Mnist Over_dim_tied_iteration 64 Mnist Over_dim_tied_iteration 32 Mnist
0 0.346407 0.334849 0.457756 0.293611 1.93055e+09 1.47257e+10 8.29947e+09 12.6991
1 0.386937 0.390018 0.455832 0.374307 9.03382e+24 7.37431e+26 4.97365e+26 358.385
2 0.407927 0.43037 0.455832 0.43774 inf inf inf 355.811
3 0.417908 0.462715 0.455832 0.469107 inf inf inf 355.811
4 0.424006 0.486917 0.455832 0.483079 inf nan nan 355.811
5 0.428975 0.500427 0.455832 0.489804 nan nan nan 355.811
6 0.432477 0.50766 0.455832 0.493094 nan nan nan 355.811
7 0.434504 0.511629 0.455832 0.494609 nan nan nan 355.811
MAE over iterations autoencoder
Over_dim_tied_iteration 256 10_Targets Over_dim_tied_iteration 128 10_Targets Over_dim_tied_iteration 64 10_Targets Over_dim_tied_iteration 32 10_Targets Over_dim_tied_iteration 256 Mnist Over_dim_tied_iteration 128 Mnist Over_dim_tied_iteration 64 Mnist Over_dim_tied_iteration 32 Mnist
0 0.272003 0.282449 0.265065 0.277778 5620 16731.4 10493 0.415634
1 0.275256 0.295306 0.265029 0.290361 3.84448e+11 3.74402e+12 2.56867e+12 1.75825
2 0.276552 0.305941 0.265029 0.30262 2.62987e+19 8.37841e+20 6.28812e+20 1.75362
3 0.277333 0.314389 0.265029 0.309511 1.79899e+27 1.87493e+29 1.53934e+29 1.75363
4 0.278053 0.320522 0.265029 0.312644 inf nan nan 1.75364
5 0.278747 0.32402 0.265029 0.314089 nan nan nan 1.75364
6 0.279206 0.326061 0.265029 0.31474 nan nan nan 1.75365
7 0.279488 0.327278 0.265029 0.315006 nan nan nan 1.75365 
predictions_df_90
Accuracy over iterations evaluations_feature_classifier
Over_dim_tied_iteration 256 10_Targets Over_dim_tied_iteration 128 10_Targets Over_dim_tied_iteration 64 10_Targets Over_dim_tied_iteration 32 10_Targets Over_dim_tied_iteration 256 Mnist Over_dim_tied_iteration 128 Mnist Over_dim_tied_iteration 64 Mnist Over_dim_tied_iteration 32 Mnist
0 0.4997 0.529 0.144 0.3873 0.4706 0.4362 0.4696 0.0932
1 0.4694 0.4499 0.1135 0.3341 0.4855 0.4583 0.4932 0.1127
2 0.4377 0.4 0.1135 0.3112 0.4775 0.4567 0.4906 0.1054
3 0.4262 0.3725 0.1135 0.2996 0.4645 0.4456 0.4802 0.1028
4 0.4175 0.3611 0.1135 0.297 0.447 0.4349 0.464 0.1029
5 0.4102 0.3548 0.1135 0.2954 0.434 0.4211 0.4474 0.1029
6 0.4049 0.3519 0.1135 0.295 0.4165 0.4107 0.4343 0.1029
7 0.4006 0.3499 0.1135 0.2948 0.3967 0.3959 0.4222 0.1029
Loss over iterations autoencoder
Over_dim_tied_iteration 256 10_Targets Over_dim_tied_iteration 128 10_Targets Over_dim_tied_iteration 64 10_Targets Over_dim_tied_iteration 32 10_Targets Over_dim_tied_iteration 256 Mnist Over_dim_tied_iteration 128 Mnist Over_dim_tied_iteration 64 Mnist Over_dim_tied_iteration 32 Mnist
0 0.350723 0.33853 0.458633 0.300344 2.7256e+09 1.93238e+10 9.3675e+09 31.7281
1 0.393889 0.398521 0.455832 0.388426 1.27542e+25 9.67693e+26 5.6137e+26 848.224
2 0.415655 0.443372 0.455832 0.457853 inf inf inf 850.991
3 0.425754 0.478088 0.455832 0.490306 inf inf inf 850.99
4 0.431976 0.503624 0.455832 0.504277 inf nan nan 850.99
5 0.436801 0.517716 0.455832 0.511011 nan nan nan 850.99
6 0.440411 0.524611 0.455832 0.514405 nan nan nan 850.99
7 0.442905 0.528114 0.455832 0.516103 nan nan nan 850.99
MAE over iterations autoencoder
Over_dim_tied_iteration 256 10_Targets Over_dim_tied_iteration 128 10_Targets Over_dim_tied_iteration 64 10_Targets Over_dim_tied_iteration 32 10_Targets Over_dim_tied_iteration 256 Mnist Over_dim_tied_iteration 128 Mnist Over_dim_tied_iteration 64 Mnist Over_dim_tied_iteration 32 Mnist
0 0.27714 0.288954 0.265105 0.284686 8250.8 22807.6 11522.1 0.641987
1 0.279903 0.302578 0.265029 0.299639 5.6442e+11 5.10372e+12 2.82059e+12 3.82172
2 0.280838 0.314086 0.265029 0.313184 3.86098e+19 1.14211e+21 6.90482e+20 3.82823
3 0.281534 0.322985 0.265029 0.320241 2.64116e+27 2.55583e+29 1.6903e+29 3.82814
4 0.28225 0.329345 0.265029 0.323402 inf nan nan 3.82813
5 0.282941 0.332916 0.265029 0.324882 nan nan nan 3.82813
6 0.283448 0.334816 0.265029 0.325589 nan nan nan 3.82813
7 0.283824 0.335815 0.265029 0.325928 nan nan nan 3.82813 
predictions_df_100
Accuracy over iterations evaluations_feature_classifier
Over_dim_tied_iteration 256 10_Targets Over_dim_tied_iteration 128 10_Targets Over_dim_tied_iteration 64 10_Targets Over_dim_tied_iteration 32 10_Targets Over_dim_tied_iteration 256 Mnist Over_dim_tied_iteration 128 Mnist Over_dim_tied_iteration 64 Mnist Over_dim_tied_iteration 32 Mnist
0 0.4116 0.4452 0.1381 0.3281 0.3905 0.3498 0.3891 0.0947
1 0.3988 0.3903 0.113 0.278 0.3982 0.3637 0.404 0.1161
2 0.3788 0.3427 0.1135 0.2634 0.3874 0.3634 0.4019 0.1066
3 0.3686 0.3217 0.1135 0.2547 0.3769 0.3609 0.3973 0.1017
4 0.3604 0.3119 0.1135 0.2507 0.3602 0.3496 0.3848 0.1014
5 0.3543 0.3067 0.1135 0.2504 0.3508 0.3403 0.3759 0.1014
6 0.3494 0.3035 0.1135 0.2503 0.3332 0.3299 0.3654 0.1014
7 0.345 0.3017 0.1135 0.2502 0.3235 0.3182 0.3541 0.1014
Loss over iterations autoencoder
Over_dim_tied_iteration 256 10_Targets Over_dim_tied_iteration 128 10_Targets Over_dim_tied_iteration 64 10_Targets Over_dim_tied_iteration 32 10_Targets Over_dim_tied_iteration 256 Mnist Over_dim_tied_iteration 128 Mnist Over_dim_tied_iteration 64 Mnist Over_dim_tied_iteration 32 Mnist
0 0.356122 0.347002 0.462827 0.311531 5.14684e+09 3.58009e+10 1.45927e+10 63.5763
1 0.400606 0.412539 0.455832 0.409514 2.40842e+25 1.79283e+27 8.74503e+26 1308.47
2 0.423798 0.459436 0.455832 0.484753 inf inf inf 1318.89
3 0.434797 0.494455 0.455832 0.51849 inf inf inf 1320.79
4 0.441665 0.518924 0.455832 0.532509 inf nan nan 1320.79
5 0.446666 0.532304 0.455832 0.538773 nan nan nan 1320.79
6 0.450795 0.538465 0.455832 0.541786 nan nan nan 1320.79
7 0.453646 0.541711 0.455832 0.543212 nan nan nan 1320.79
MAE over iterations autoencoder
Over_dim_tied_iteration 256 10_Targets Over_dim_tied_iteration 128 10_Targets Over_dim_tied_iteration 64 10_Targets Over_dim_tied_iteration 32 10_Targets Over_dim_tied_iteration 256 Mnist Over_dim_tied_iteration 128 Mnist Over_dim_tied_iteration 64 Mnist Over_dim_tied_iteration 32 Mnist
0 0.284071 0.296984 0.265219 0.293099 15140.9 40761.3 19023.8 1.0109
1 0.286092 0.311423 0.265029 0.311138 1.03576e+12 9.12133e+12 4.65701e+12 5.74944
2 0.286653 0.322865 0.265029 0.32649 7.08525e+19 2.04118e+21 1.14004e+21 5.7915
3 0.287124 0.331344 0.265029 0.333993 4.84676e+27 4.56777e+29 2.79083e+29 5.79657
4 0.287837 0.337172 0.265029 0.33716 inf nan nan 5.79653
5 0.288366 0.340448 0.265029 0.338538 nan nan nan 5.79652
6 0.288888 0.342092 0.265029 0.339146 nan nan nan 5.79652
7 0.289341 0.343004 0.265029 0.339424 nan nan nan 5.79652 
/home/hicky/anaconda3/lib/python3.7/site-packages/matplotlib/image.py:397: UserWarning: Warning: converting a masked element to nan.
  dv = (np.float64(self.norm.vmax) -
/home/hicky/anaconda3/lib/python3.7/site-packages/matplotlib/image.py:398: UserWarning: Warning: converting a masked element to nan.
  np.float64(self.norm.vmin))
/home/hicky/anaconda3/lib/python3.7/site-packages/matplotlib/image.py:405: UserWarning: Warning: converting a masked element to nan.
  a_min = np.float64(newmin)
/home/hicky/anaconda3/lib/python3.7/site-packages/matplotlib/image.py:410: UserWarning: Warning: converting a masked element to nan.
  a_max = np.float64(newmax)
<string>:6: UserWarning: Warning: converting a masked element to nan.
/home/hicky/anaconda3/lib/python3.7/site-packages/numpy/ma/core.py:711: UserWarning: Warning: converting a masked element to nan.
  data = np.array(a, copy=False, subok=subok)

/home/hicky/anaconda3/lib/python3.7/site-packages/matplotlib/image.py:397: UserWarning: Warning: converting a masked element to nan.
  dv = (np.float64(self.norm.vmax) -
/home/hicky/anaconda3/lib/python3.7/site-packages/matplotlib/image.py:398: UserWarning: Warning: converting a masked element to nan.
  np.float64(self.norm.vmin))
/home/hicky/anaconda3/lib/python3.7/site-packages/matplotlib/image.py:405: UserWarning: Warning: converting a masked element to nan.
  a_min = np.float64(newmin)
/home/hicky/anaconda3/lib/python3.7/site-packages/matplotlib/image.py:410: UserWarning: Warning: converting a masked element to nan.
  a_max = np.float64(newmax)
<string>:6: UserWarning: Warning: converting a masked element to nan.
/home/hicky/anaconda3/lib/python3.7/site-packages/numpy/ma/core.py:711: UserWarning: Warning: converting a masked element to nan.
  data = np.array(a, copy=False, subok=subok)